Neural networks have proven apparent success in solving a wide variety of problems. Various engineering applications in particular were studied using these powerful tools.

The Neural network resembles the function of the human brain. A training

process has to be applied first in order that the network acquires knowledge through learning. Then the network can generalize predicted solutions based on the a mount and accuracy of the training data. The target of this research is to develop two neural networks that can predict the productivity of two common construction-finishing activities, which are brick and plaster works. In order to adequately train the networks, 12 projects were used to gather 85 and 100 data points for the brick works and plaster works respectively. These construction projects included residential, commercial, and industrial projects. The gathered data targeted measuring the factors affecting the productivity as well as the related productivity values. Several trials were conducted to choose the appropriate network architecture. Finally 85% of the data was used to train the networks and 15% was used for testing the generalization capability of the networks. The results showing the desired and output values for the testing data were discussed and analyzed. Furthermore, the training data was used again to validate the generalization capability of the networks. The results were also discussed and compared for a predetermined accuracy level. A sensitivity analysis was performed to relate the effects of each of the contributing factors on the productivity values.


School of Sciences and Engineering


Interdisciplinary Engineering Program

Date of Award


Online Submission Date


First Advisor

Ahmed Samer Ezeldin

Committee Member 1

Ibrahim Abdel Rashid

Committee Member 2

Ahmed Sherif

Committee Member 3

Medhat Haroun

Document Type



201 leaves

Library of Congress Subject Heading 1

Neural networks (Computer science)

Library of Congress Subject Heading 2



The American University in Cairo grants authors of theses and dissertations a maximum embargo period of two years from the date of submission, upon request. After the embargo elapses, these documents are made available publicly. If you are the author of this thesis or dissertation, and would like to request an exceptional extension of the embargo period, please write to thesisadmin@aucegypt.edu

Call Number

Thesis 2003/55